The Last Fossil Primate in North America, New Material of the Enigmatic Ekgmowechashala from the Arikareean of Oregon Joshua X

Home , Adapiformes, Cantius, Herpetotherium, John Day River, Nanotragulus, Smilodectes

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

U.S. National Park Service Publications and Papers National Park Service

2015 The Last Fossil Primate in North America, New Material of the Enigmatic Ekgmowechashala From the Arikareean of Oregon Joshua X. Samuels National Park Service

L. Barry Albright University of North Florida

Theodore J. Fremd National Park Service

Follow this and additional works at: http://digitalcommons.unl.edu/natlpark

Samuels, Joshua X.; Albright, L. Barry; and Fremd, Theodore J., "The Last Fossil Primate in North America, New Material of the Enigmatic Ekgmowechashala From the Arikareean of Oregon" (2015). U.S. National Park Service Publications and Papers. 135. http://digitalcommons.unl.edu/natlpark/135

This Article is brought to you for free and open access by the National Park Service at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in U.S. National Park Service Publications and Papers by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 158:43–54 (2015)

The Last Fossil Primate in North America, New Material of the Enigmatic Ekgmowechashala From the Arikareean of Oregon

Joshua X. Samuels,1* L. Barry Albright,2 and Theodore J. Fremd1,3

1National Park Service, John Day Fossil Beds National Monument, Kimberly, OR 97848 2Department of Physics, University of North Florida, Jacksonville, FL 32224 3Department of Geological Sciences, Museum of Natural and Cultural History, University of Oregon, Eugene, OR 97403

KEY WORDS Ekgmowechashalinae; Adapiformes; John Day Formation

ABSTRACT OBJECTIVE: Primates were common in chashala to a pair of recently described Asian primates, North America through most of the Eocene, but vanished in Muangthanhinius and Bugtilemur, suggests that it is a the Chadronian, about 35 million years ago. In the Arikar- strepsirrhine adapiform, rather than an omomyid. The eean, about 6 million years later, the enigmatic primate well-deﬁned stratigraphy and dated marker beds of the Ekgmowechashala appeared in the Great Plains and Ore- Turtle Cove Member provide a reﬁned age for Ekgmowe- gon. This taxon shows little resemblance to other North chashala occurrences in Oregon, during the Oligocene American primates and its phylogenetic position has long (early Arikareean). been debated. New material of this taxon allows a revised CONCLUSIONS: The age and morphology of these ekg- assessment of its age and how it is related to other primates. mowechashaline taxa suggest that the group originated METHODS: Recently collected Ekgmowechashala speci- in Asia and dispersed to North America in the Oligo- mens from the Turtle Cove Member of the John Day cene, after the extinction of other primates in North Formation in Oregon are described. These specimens are America. Contemporaneous occurrences of Ekgmowecha- compared to previously collected material from South shala in Oregon and the Great Plains indicate the last Dakota and Nebraska, as well as other fossil primates non-human primates vanished in North America about from North America and Asia. 26 million years ago. Am J Phys Anthropol 158:43–54, RESULTS: Study of the John Day material allows diag- 2015. VC 2015 Wiley Periodicals, Inc. nosis of a new, distinct species. Comparison of Ekgmowe-

During the Eocene, the earth was in a ‘greenhouse’ The phylogenetic position of this taxon has been long state free from permanent ice sheets (Zachos et al., 2001). debated, with various authors placing Ekgmowechashala Euprimates ﬁrst appeared in North America during the in the Omomyidae (Primates) (MacDonald, 1963, 1970; early Eocene (Bown and Rose, 1987; Smith et al., 2006; Szalay, 1976; Szalay and Delson, 1979; Rose and Beard, 2008) and much of North America was inhabited Rensberger, 1983; Gunnell et al., 2008), Plagiomenidae by primates through the middle Eocene (Gunnell et al., (Dermoptera) (McKenna, 1990), and most recently in the 2008). During the late middle Eocene (42–37 Ma) global Adapiformes (Primates) (Ni et al., 2010). climate conditions changed dramatically. In North Amer- John Day Formation strata in the Sheep Rock area (Fig. ica, this resulted in a shift from semi-tropical closed can- 1) have been studied for about 150 years. The best known opy forests to more open woodland and savannah assemblage from that area is the Oligocene aged Turtle habitats, and a corresponding decrease in primate diver- Cove fauna, which includes nearly 100 recognized mammal sity (Storer, 1990; Janis, 1993; Wing, 1998; Gunnell et al., species (Fremd, 2010). In 1997, the ﬁrst stratigraphically 2008; Townsend et al., 2010). Multiple primate taxa sur- assignable specimen of Ekgmowechashala,aleftm1,was vived into the Duchesnean, but only the omomyid Roo- neyia from Texas and Chumashius survived into the early Chadronian (late Eocene) (Wilson, 1966; Ostrander, 1987). Additional Supporting Information may be found in the online After the extinction of Rooneyia and Chumashius almost version of this article. 35 million years ago, primates were absent from North America for about 6 million years, until the appearance of This article first appeared online on 29 June 2015 and has since been the enigmatic Ekgmowechashala in the early Arikareean changed. (MacDonald, 1963, 1970; Rose and Rensberger, 1983). Ekgmowechashala was originally known from nine *Correspondence to: Joshua Samuels, National Park Service, John Day Fossil Beds National Monument, Kimberly, OR 97848, dentary specimens from the Upper Sharps Formation of USA. E-mail: [email protected] South Dakota (MacDonald, 1963, 1970; Szalay, 1976), but a maxilla fragment with teeth from the John Day Received 15 January 2015; revised 1 May 2015; accepted 6 May Formation of Oregon was later assigned to this taxon 2015 (Rose and Rensberger, 1983). This rare primate has also been noted from the contemporaneous Gering Formation DOI: 10.1002/ajpa.22769 of Nebraska and Turtle Butte Formation of South Published online 29 June 2015 in Wiley Online Library Dakota (Swisher, 1982; Schumaker, 2008; Martin, 2011). (wileyonlinelibrary.com).

Ó 2015 WILEY PERIODICALS, INC. 44 J.X. SAMUELS ET AL. recovered from Turtle Cove unit H, a stratum containing early Arikareean (Ar1) fauna. After more than a decade of diligent searching, the exact same site yielded two additional specimens, a p4 and m3, in 2011. And in early 2015 a dentary fragment that fits the p4 and m1 was discovered, confirming that these teeth are from the same individual. In 2013 and 2014, two other partial m1s were collected from a second site on Sheep Rock, also within unit H. The well-defined stratigraphy and a series of dated marker beds at these sites allow us to place these new specimens of Ekgmowechashala in a chronological and biostratigraphic context. Study of these new specimens confirms the presence of Ekgmowechashala in Oregon, facilitates direct comparison to specimens from the Great Plains, and refines the age of occurrence of the last non-human primate in North America.

MATERIALS AND METHODS The specimens newly described here are all housed in Fig. 1. Map showing the Sheep Rock Unit of John Day the collection of John Day Fossil Beds National Monu- Fossil Beds National Monument, Grant County, Oregon. The ment. Comparisons made in this article are based pri- asterisk indicates the location of Sheep Rock, where the specimens described here were found. marily on study of casts of fossils at other institutions, supplemented by published photographs and drawings. Dental nomenclature used in comparative descriptions middle Eocene to early Miocene in age, about 39 to 18 follows that of Szalay (1976) and a labeled illustration is Ma (Hunt and Stepleton, 2004; Albright et al., 2008). included as a Supporting Information figure (Fig. S1). All of the specimens described here originate from Upper teeth are designated by capital letters and lower within the Turtle Cove Member of the John Day Forma- teeth by lower-case letters (e.g., M1 and m1). Measure- tion (Fig. 2), which is characterized by lithology consist- ments represent maximum lengths and widths taken at ing of zeolitized bluish-green and tan claystones and the occlusal surface. These measurements were taken siltstones (Fisher and Rensberger, 1972). Fremd et al. using Mitutoyo Absolute digital calipers to the nearest (1994) subdivided the Turtle Cove Member and the over- 0.01 mm. Precise locality information for recent collec- lying Kimberly Member into lithostratigraphic units A- tions, including GPS (global positioning satellite) data, is M, based on the individual lithologies of beds and inter- on file at John Day Fossil Beds National Monument. spersed tuffs. There are currently 12 radiometric dates Abbreviations for institutions: JODA, John Day Fossil from the John Day Formation, including six tuffs within Beds National Monument; LACM, Natural History the Turtle Cove Member alone. These tuffs were radio- Museum of Los Angeles County; SDSM, South Dakota metrically dated using 40Ar/39Ar single-crystal laser- School of Mines and Technology, Museum of Geology; fusion (Albright et al., 2008). The radiometric dates rele- UCMP, University of California Museum of Paleontology. vant to this study are the following: Picture Gorge Ignimbrite (Member H of Robinson) dated 28.7 6 0.07 GEOLOGICAL SETTING Ma, Deep Creek Tuff dated 27.89 6 0.57 Ma, Biotite Tuff dated 27.14 6 0.13 Ma, and Tin Roof Tuff dated Distributed widely throughout central and eastern Ore- 25.9 6 0.3 Ma (Albright et al., 2008). gon, the John Day Formation includes 1000 m of geo- graphically variable strata, consisting mainly of SYSTEMATIC PALEONTOLOGY volcaniclastic sedimentary rocks and airfall tuffs (Fisher Order Primates Linnaeus, 1758 and Rensberger, 1972; Robinson et al., 1984; Retallack Semiorder Euprimates Hoffstetter, 1977 et al., 2000; Albright et al., 2008, McClaughry et al., 2009). Suborder Strepsirrhini Geoffroy Saint-Hilaire, 1812 These strata are well understood after extensive study for Infraorder Adapiformes Hoffstetter, 1977 more than 100 years (Merriam, 1901; Hay, 1963; Fisher Family Incertae Sedis and Rensberger, 1972; Fremd et al., 1994; Hunt and Steple- Subfamily Ekgmowechashalinae Szalay, 1976 ton, 2004; Albright et al., 2008). A series of conspicuous Genus Ekgmowechashala Macdonald, 1963 marker units and distinctive lithologies of layers throughout the sequence have allowed exposures in Oregon to be Type species correlated at great distances (Albright et al., 2008). Three distinct “facies” of the John Day Formation Ekgmowechashala philotau have been described: western, southern, and eastern (Robinson et al., 1984). A nearly continuous sequence of Included species strata is exposed within outcrops of the eastern “facies,” Ekgmowechashala zancanellai, new species preserved in the vicinity of the Sheep Rock Unit of John Day Fossil Beds National Monument. The recent work Emended diagnosis of Hunt and Stepleton (2004) and Albright et al. (2008) have produced a detailed litho- and chronostratigraphy Teeth are low crowned with bulbous cusps and weakly for the John Day Formation, which is paired with radioi- developed cristae; upper and lower molars decrease in sotopic and paleomagnetic calibration (Albright et al., size distally; P4 and p4 molarized; double-rooted p2; p4 2008: figs. 9 and 10). As currently recognized, the John with enlarged talonid, metastylid, ectostylid, and several Day Formation includes seven members, ranging from small cuspules; trigonids of lower molars mesiodistally

American Journal of Physical Anthropology LAST FOSSIL PRIMATE IN NORTH AMERICA 45

Fig. 2. Composite stratigraphic section of the John Day Formation near Sheep Rock with stratigraphic position of the known Ekgmowechashala specimens. Stratigraphy and radiometric dates are based on Albright et al. (2008) and Fremd (2010). The black bar indicates conﬁrmed or inferred stratigraphic ranges; grey bar indicates range-through. [Color ﬁgure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] compressed; paraconids absent and paracristids weak to Horizon and locality absent in lower molars; lower molar hypoconulids distinct, twinned with hypoconid; P4 with distally displaced JODA 6322, 16525, 16526 – JDNM 8, Sheep Rock, paracone and large paraconule, located mesial to and Grant County, OR, Unit H, Turtle Cove Member, John between protocone and paracone; prominent parastyle Day Formation; UCMP 128231 – UCMP V82379, Grant present in P4 and M1; M1 and M2 with large pseudohy- County, OR, unknown unit, upper Turtle Cove Member, pocone arising from the “Nannopithex-fold” of the proto- John Day Formation. cone; tooth enamel crenulated. Distribution Age Early Arikareean (early? and late Oligocene, Ar1 and Early Arikareean (early? and late Oligocene, Ar1 and Ar2) of South Dakota, Nebraska, and Oregon. Ar2), between Picture Gorge Ignimbrite dated New species 28.7 6 0.07 Ma and Tin Roof Tuff dated 25.9 6 0.3 Ma (Albright et al., 2008). Ekgmowechashala zancanellai new species (Figs. 3–6)

Type specimen Diagnosis JODA 6322, left dentary fragment with p4, m1, and p4 paracristid more distinct than in E. philotau and m3 (Figs. 3–4). located just lingual to midline of tooth; proportions of Referred specimens lower dentition distinct from E. philotau, m1 substantially enlarged and m3 reduced relative to other cheek UCMP 128231, left maxilla with P4, M1, and partial M2; teeth; m3 paracristid distinct and broad, enclosing a JODA 16525, partial left m1; JODA 16526, partial left m1. small trigonid basin.

American Journal of Physical Anthropology 46 J.X. SAMUELS ET AL.

Fig. 5. Comparison of lower dentition among known ekgmowechashaline primates. A. Ekgmowechashala zancanellai,JODA 6322, left p4, m1, and m3; B. E. philotau, composite lower dentition (p2 to m3) based on SDSM 5550, SDSM 61204, and LACM 9207; C. Bugtilemur mathesonae, UMC-DBC 2139, left p3, UMC- DBC 2143, left p4, UMC-DBC 2141, left m1, UMC-DBC 2161, left m2; D. Muangthanhinius siami, TF 6212, left dentary with p4 to m1. Scale bars equal 2 mm. Illustrations adapted from the following sources: E. philotau, Szalay (1976:Fig. 133); B. mathesonae copyright VC by Laurence Meslin-CNRS, Marivaux et al. (2001, fig. 2); and M. siami VC by Laurence Meslin-CNRS, Marivaux et al. Fig. 3. Ekgmowechashala specimens from the Arikareean of (2006, fig. 3). Reproduced by permission of Laurence Meslin. Origi- Oregon. A. JODA 6322, holotype of Ekgmowechashala zancanel- nal illustration of E. zancanellai by Keila Bredehoeft. VC Repro- lai, left p4, m1, and m3; B. JODA 16525, partial left m1; C. duced by permission of Laurence Meslin. JODA 16526, partial left m1; D. grayscale cast of UCMP 128231, left maxilla with P4, M1, and partial M2. Scale bar equal 2 mm. [Color figure can be viewed in the online issue, Etymology which is available at wileyonlinelibrary.com.] Patronym for Bureau of Land Management archeolo- gist John Zancanella, who discovered the type specimen and whose dedicated paleontological efforts helped to discover and preserve many important localities in central and eastern Oregon. DESCRIPTION While represented by limited material, the lower teeth of the holotype of Ekgmowechashala zancanellai (JODA 6322, Fig. 3) are complete and well-preserved enough to distinguish them from E. philotau. As is characteristic of the genus, the teeth display low, bulbous cusps and weakly developed cristae, with well-developed irregular cingulids. The enamel of both the p4 and molars is crenulated, particularly within the talonid basin and on the surfaces of some cusps. The p4 of JODA 6322 is semimolariform and nearly rectangular in shape with a low, irregular cingulid sur- rounding the mesial and buccal borders of the tooth. There is no paraconid, but a small, distinct paracristid is present just lingual to the midline of the tooth, mesial to and between the protoconid and metaconid. A small trigonid basin lies distal to the paracristid, formed primarily by the relatively large buccally located protoconid and Fig. 4. Occlusion of upper and lower dentitions of Ekgmowe- mesiolingually located metaconid. Distobuccal to the pro- chashala zancanellai from Oregon, in lateral view. Illustrated above is the maxillary dentition of UCMP 128231 (grayscale toconid is a large ectostylid (ectostylar cusp of Rose and cast) and below is the mandibular dentition of JODA 6322, both Rensberger, 1983), nearly equal in size to the protoconid. in lateral view. Scale bar equals 2 mm. [Color figure can be Distal and lingual to this stylid is the hypoconid, which viewed in the online issue, which is available at wileyonlineli- has a pair of crests extending from it. One crest extends brary.com.] buccally from the hypoconid and connects to the distal

American Journal of Physical Anthropology LAST FOSSIL PRIMATE IN NORTH AMERICA 47 the protoconid and the hypoconid. The hypoconid is crescentic in shape and bears a very weakly developed cristid obliqua, which runs mesially from the hypoconid along the buccal margin of the tooth. Distolingual to the hypoconid is a smaller, closely apressed hypoconulid. Extending lingually from the hypoconulid is the postcristid, which extends and connects to the distolingual corner of the entoconid. A prominent irregular cingulid extends from the mesiobuccal edge of the protoconid along the buccal margin of the m1 and wraps around the distal surface of the hypoconid. Two partial m1s are included in the JODA collection, JODA 16525 and 16526. Both preserve only the distal part of the tooth, including the entoconid, hypoconid, hypoconulid, and part of the talonid basin. The talonid basins of both specimens are distinctly crenulated, as is characteristic of Ekgmowechashala (Macdonald, 1963; Gunnell et al., 2008). These teeth are identified as m1 rather than m2 based on the fact that the hypoconulid is relatively prominent and distinct from the hypoconid, while the m2 hypoconulid is simply a small projection from the distal cingulid of the hypoconid (McKenna, 1990). Both specimens are substantially smaller than JODA 6322, with JODA 16525 10% and JODA 16526 nearly 13% smaller (Table 1). The m3 of JODA 6322 is elongate, with a rather dis- Fig. 6. Comparison of upper dentition among known ekgmo- tinctive arrangement of cusps. As in the m1 and m2, wechashaline primates. A. Ekgmowechashala zancanellai,UCMP there is no paraconid, though the paracristid is much 128231, left maxilla with P4, M1, and partial M2; B. Bugtilemur mathesonae, UMC-DBC 2147, partial right P4, UMC-DBC 2151, more distinct and substantial than in the other molars, left M1, UMC-DBC 2146, right M2, and UMC-DBC 2152, right running transversely across the entire mesial end of the M2. Scale bars equal 2 mm. Illustration of B. mathesonae copy- m3. There is a mesiodistally narrow, transversely ori- right VC by Laurence Meslin-CNRS adapted from Marivaux et al. ented trigonid basin between the paracristid and the pro- (2001, fig. 2). Reproduced by permission of Laurence Meslin. toconid and metaconid. The protoconid and metaconid are Original illustration of E. zancanellai by Keila Bredehoeft. nearly equal in size and are connected via a much more VC Reproduced by permission of Laurence Meslin. prominent protocristid than is present in the m1. A small crest extends distally from the protoconid; this crest bears a small cuspule as in the m1. The metaconid is mesiodis- surface of a small, but distinct cuspule in the distobuccal tally compressed and has a distinct buccally directed crest corner of the tooth; the other crest (postcristid) extends that joins the protocristid. The metastylid is smaller than lingually from the hypoconid to the distolingual corner the metaconid and just distal to it, and well separated of the tooth. Along the buccal cingulid between the disto- from the similarly sized entoconid. The hypoconid is large buccal cuspule and the ectostylid is another small cingu- and located more mesially than in the anterior molars, lar cuspule. Distal to the metaconid is a relatively large mesial to the entoconid, yielding a substantially smaller metastylid, which is well-separated from both the meta- talonid basin. A low crest runs distolingually from the conid and entoconid. Just distal to the metastylid on the hypoconid to the entoconid. The hypoconid is separated lingual border of the p4 is a much smaller entoconid. distally from the similarly large hypoconulid. Extending Also located on the lingual border of the p4 distal to the from the hypoconulid across the distal surface of the m3 entoconid and mesiolingual to the lingual extension of is a broad, crescent-shaped postcristid. The postcristid the postcristid is an additional small neomorphic cus- extends to the distolingual corner of the tooth, just distal pule, nearly equal in size to the entoconid. to the entoconid. The distal part of the m3 is relatively The m1 of JODA 6322 is rectangular in occlusal out- worn, resulting in large concavities in the surfaces of the line, with an enlarged talonid basin. As in the p4, the hypoconid and hypoconulid. m1 has no paraconid. The mesial end of the tooth has a Upper dentition of UCMP 128231 (Fig. 3) was previ- large buccal protoconid and lingual metaconid; the proto- ously described in detail by Rose and Rensberger (1983), cristid between these cusps is rather poorly developed. A and the specimen was referred to Ekgmowechashala sp. flat precingulid runs along the surface mesial to the pro- based on details of dental morphology and proper occlu- toconid and metaconid, possibly representing a remnant sion of the specimen with a specimen of Ekgmowecha- of the paracristid. The protoconid has a small mesial shala philotau (SDSM 62104) from South Dakota. The crest that may be a remnant of the paracristid, as well description that follows is based mainly on the descrip- as a distal crest that extends distolingually from its dis- tion of Rose and Rensberger (1983), with some modifica- tobuccal corner, and which bears a pair of small cus- tions based on more recent studies. The teeth in this pules. The metaconid is mesiodistally compressed and specimen are characterized by bulbous cusps, weakly nearly equal in size to the protoconid. Distal to the developed cristae, and well-developed irregular cingula metaconid is a large, bulbous metastylid. Distal to the wrapping around the teeth. Crenulated enamel is evi- metastylid is a similarly sized entoconid, nearly located dent within basins and on unworn surfaces of cusps. in the distolingual corner of the tooth. The talonid basin The upper P4 of UCMP 128231 is trapezoidal in shape is enlarged and open buccally between the distal crest of and semimolariform. The tooth has a large, low

American Journal of Physical Anthropology 48 J.X. SAMUELS ET AL.

TABLE 1. Measurements (in mm) of Ekgmowechashala specimens, some data derived from Macdonald (1963, 1970) Ekgmowechashala zancanellai Turtle Ekgmowechashala philotau Wounded Knee fauna, Cove Sharps Fm. South Dakota fauna, John Day Fm. Oregon SDSM SDSM SDSM SDSM LACM LACM JODA JODA JODA UCMP 5550 5551 55111 62104 9206 9207 6322 16525 16526 128231 p4L 3.3 3.0 3.20 3.49 3.67 p4W 2.8 2.7 2.73 3.11 3.08 m1L 3.6 3.5 4.2 3.72 4.17 4.06 4.46 m1W 3.3 3.3 3.5 3.55 3.35 3.48 3.78 3.41 3.30 m2L 3.4 3.3 3.9 3.38 m2W 2.8 2.8 2.8 3.09 m3L 3.25 3.21 m3W 2.42 2.29 P4L 3.35 P4W 4.20 M1L 3.58 M1W 5.21 M2L 3.34a M2W 4.87a p4L/m1L 0.917 0.857 0.860 0.860 0.823 m3L/m1L 0.874 0.720 a Approximate based on incomplete teeth and alveoli. protocone in the mesiolingual corner, no hypocone, and closely appressed but distinct. The enamel of the basin the paracone and metacone are closely appressed. The buccal to the protocone and pseudohypocone is heavily paracone is larger and higher than the metacone, and crenulated. lies near the midpoint of the buccal margin. Between and As indicated by Rose and Rensberger (1983), the dis- slightly mesial to the protocone and paracone is a promi- tally displaced paracone of Ekgmowechashala is unique nent paraconule, which is larger but lower than the meta- among primates. The paracone of the P4 occludes cone. A distinct parastyle lies in the mesiobuccal corner mesially with the distal surface of the p4 protoconid and of the tooth. The tooth is surrounded by a low, well devel- lingual surface of the p4 ectostylid (Fig. 4). Similarly, oped cingulum; the mesial part of the cingulum is broken, the M1 paracone occludes mesially with the distal sur- while the distal part is broadened and includes several face of the m1 protoconid and mesial surface of the hypo- tiny worn cuspules. The major cusps (protocone, paraco- conid (Fig. 4). Worn surfaces of cusps have exposed nule, paracone, and metacone) form a semicircle, with a dentine; this wear is more substantial in the lingual broad posterointernal basin distal to them. cusps of the upper teeth. Comparison of the M2 to P4 The M1 of UCMP 128231 is larger and more quadrate and M1 reveals enamel crenulation in the upper cheek than the P4. As in the P4, the paracone is near the cen- teeth was substantial, though reduced with wear. Simi- ter of the buccal margin of the tooth, with a distinct par- lar comparison of several m1 specimens shows extensive astyle mesial and closely appressed metacone distal to crenulation is present in the talonid basin and surfaces it. The paraconule and metaconule are low and large, of cusps in unworn teeth (JODA 16526), but crenulation lying mesiolingual to the paracone and metacone respec- was reduced with wear (JODA 6322 and 16525) (Fig. 3). tively. The crescentic protocone lies in the mesiolingual corner of the tooth, with an oblique mesiobuccal to disto- COMPARISONS lingual orientation. A narrow crest extends buccally from the paracone, ending adjacent to the parastyle and The dentition of Ekgmowechashala is unusual among not connecting to the paraconule. Connected to and primates. In both the upper and lower dentitions, the extending from the distal edge of the protocone is a molars decrease in size posteriorly. The P4 is semimo- prominent mesiodistally oriented “pseudohypocone.” lariform with a distinct metacone. The upper molars are This cusp is called a “pseudohypocone” due to the fact characterized by bulbous cusps and well developed that it arises from the “Nannopithex-fold” of the proto- basins, and bear a prominent “pseudohypocone.” The p4 cone, rather than from the postcingulum (Ni et al., 2010; is semimolariform with a prominent ectostylid between Anemone et al., 2012). The trigon is closed mesially, but and buccal to the protoconid and hypoconid. The lower is open distally between the ‘pseudohypocone’ and meta- molars are broad and low crowned, the paraconid is conule. Like the P4, the M1 is surrounded by a low, well absent and a large metastylid present, and the talonid developed cingulum. This cingulum is broad and shelf- basin is enlarged and shallow. Crests in the upper and like distally, with several worn cuspules, and the lingual lower molars, including the cristid obliqua, are poorly surface, particularly adjacent to the protocone, is irregu- deﬁned, suggesting that shearing was not well devel- lar and “vaguely beaded” (Rose and Rensberger, 1983). oped. In both the upper and lower dentition of Ekgmowe- The M2 of UCMP 128231 is incomplete, with the buc- chashala, the enamel is crenulated. cal half missing. The preserved lingual portion of the The John Day Ekgmowechashala specimens show M2 is mesiodistally narrower than the M1, and the some morphological and proportional differences from alveoli preserved in the maxilla indicate the tooth was previously described material of E. philotau (Fig. 5, also transversely narrower. The M2 protocone and pseu- Table 1). While the range of variation in that rare taxon dohypocone are similar in shape to the M1. These cusps is unknown, the Oregon specimens are distinct are less worn than in the M1, revealing the cusps are from material from the Great Plains. The p4 paracristid

American Journal of Physical Anthropology LAST FOSSIL PRIMATE IN NORTH AMERICA 49 of E. zancanellai is prominent and much larger than adapiforms (Gingerich, 1995; Gunnell et al., 2008). The that of E. philotau, where it is nearly absent. All molars double-rooted p2 observed in Ekgmowechashala is only in E. philotau show a highly reduced paracristid, merely observed in some adapiforms and stem primates, and present as a small precingulid mesial to the protoconid seen in no known haplorhines (Ni et al., 2010). Similarly, and metaconid. In E. zancanellai, the m3 paracristid is the distal premolars of many adapiforms are semimolari- a distinct, broad ridge running across the entire mesial form and both the p4 and P4 of Ekgmowechashala are end of the tooth, yielding a much larger m3 trigonid molarized. As is the case for Ekgmowechashala, in ada- than is present in E. philotau. The molars of Ekgmowe- piforms the talonid basins of the lower molars are typi- chashala decrease in size posteriorly, and the propor- cally broad and enamel is often crenulated (Gunnell tions of the molars of E. zancanellai are distinct from E. et al., 2008), though it is worth noting this is also the philotau. While the size of the p4 is relatively similar in case for large late occurring omomyines like Macrotar- both species, the m1 is enlarged and the m3 is dramati- sius, Ourayia, Mytonius, and Chipetaia. The m3 of Ekg- cally reduced in E. zancanellai (Table 1); in SDSM mowechashala has a particularly broad third lobe, which 61204 the m3 is 87% the length of the m1, but in JODA is often the case among adapiforms (Gunnell et al., 6322 the m3 is only 72% the length of the m1. 2008). As pointed out by Ni et al. (2010), the large The new specimens of Ekgmowechashala from Oregon, “pseudohypocone” in the upper molars of Ekgmowecha- as well as some recent discoveries from Asia, facilitate shala arises from the “Nannopithex-fold” (postprotocing- re-evaluation of this species’ relationships to other pri- ulum) of the protocone, which is found in some mates. Despite the unusual morphology of Ekgmowecha- adapiforms (Simons, 1962; Rose, 2006). shala, it has traditionally been placed within the Two recently described primates from Asia, Bugtile- Omomyidae (Macdonald, 1963, 1970; Szalay, 1976; Sza- mur mathesoni and Muangthanhinius siami (Marivaux lay and Delson, 1979; Rose and Rensberger, 1983; Szalay et al., 2001, 2006; Figs. 5–6), are similar to Ekgmowe- and Lucas, 1996; Albright, 2005; Gunnell et al., 2008). chashala and were recently assigned to the Ekgmowe- This placement was based on the presence of lower chashalinae (Seiffert, 2007). Both Bugtilemur and molars with the hypoconulid displaced toward the hypo- Muangthanhinius have a molarized p4, reduced or conid, and upper molars with a well-developed absent p4 paraconid and hypoconulid, mesiodistally com- “hypocone” (identiﬁed as a ‘pseudohypocone’ by Ni et al., pressed molar trigonids, reduced or absent molar paraco- 2010) and conules located in the buccal half of the teeth. nids, distinct molar hypoconulids and entoconids, and Szalay (1976) suggested that the origin of Ekgmowe- lateral cristid obliqua (Fig. 5). All three taxa have crenu- chashala may be in an omomyid like Rooneyia viejaen- lated enamel in basins, but the crenulation in Ekgmowe- sis. Like Ekgmowechashala, Rooneyia has cheek teeth chashala is more substantial, covering much of the tooth with bulbous cusps. However, in Rooneyia the P4 meta- surface. Like Ekgmowechashala, Muangthanhinius has cone is absent, as is typical of most omomyids other a large canine and double-rooted p2, though no canine than Shoshonius (Szalay, 1976). Shoshonius lacks the P4 or p2 specimens have been reported of Bugtilemur (Mar- paraconule, and it also has a small hypocone and a large ivaux et al., 2001, 2006). Like Ekgmowechashala, Bugti- mesostyle in the upper molars, unlike Ekgmowecha- lemur has broad talonid basins and widely separated shala. The prominent hypocone of the upper molars in protoconids and metaconids (Marivaux et al., 2001). Also Rooneyia and Washakius arises from the distal cingu- as like Ekgmowechashala, the P4 in Bugtilemur is lum, unlike that of Ekgmowechashala (McKenna, 1990; molarized, with a large crescentic protocone and well- Ni et al., 2010). developed metacone, and the M1 and M2 bear a distinct Noting differences from omomyids, McKenna (1990) parastyle and distally open trigon (Marivaux et al., reevaluated Ekgmowechashala and referred it to the 2001; Fig. 6). Plagiomenidae. Various problems with this assessment The cuspids of the lower cheek teeth are bulbous and were indicated by Szalay and Lucas (1996), as well as low crowned in Ekgmowechashala and Muangthanhi- Gunnell et al. (2008) and Ni et al. (2010). While relation nius, but more crest-like in Bugtilemur (Marivaux et al., to plagiomenids is not supported, the dissimilarities of 2001, 2006; Fig. 5). While neither of these Asian taxa Ekgmowechashala to known omomyids noted by displays an ectostylid distobuccal to the protoconid, they McKenna (1990) are valid. Ekgmowechashala has a do have a distinct, laterally positioned postprotoconid number of features that clearly separate it from omo- ridge. In both Muangthanhinius and Ekgmowechashala, myids, some of which have long been recognized (Szalay, there are prominent buccal cingulids on the lower p4 1976; McKenna, 1990). Unlike known omomyids, Ekg- and m1. The known material of Muangthanhinius sug- mowechashala has: 1) molars that decrease in size dis- gests that the lower molars decreased in size posteriorly tally, 2) molarized fourth premolars (p4 and P4), 3) (Marivaux et al., 2006), as is characteristic of Ekgmowe- enlarged canines, 4) double-rooted p2, 5) paraconids of chashala.InBugtilemur, the m2 is larger than the m1, lower molars absent, 6) hypoconulids of lower molars suggesting the molars did not decrease in size posteri- distinct, rather than a mere bulge in the postcristid, 7) orly (Marivaux et al., 2001). Unlike Ekgmowechashala, large bulbous parastyles of upper P4 and M1, and 8) the M1 and M2 of Bugtilemur lack a “pseudohypocone” large “pseudohypocone” in M1 and M2 arising from the and conules (Marivaux et al., 2001; Fig. 6). “Nannopithex-fold” of the protocone rather than the dis- These two Asian ekgmowechashaline taxa have more tal cingulum (Ni et al., 2010). distinct paracristids and larger trigonid basins than The combination of characters present in Ekgmowe- Ekgmowechashala (Fig. 5), though these features are chashala suggests that it is not an omomyid, and several not nearly as large as in most other adapiforms. Addi- recent studies have placed it among the adapiforms tionally, the p4 and m1 of these taxa lack distinct meta- (Marivaux et al., 2006; Seiffert, 2007; Ni et al., 2010). As stylids, but instead show relatively strong evidenced by SDSM 62104 and LACM 9207, the canine postmetacristids. In both Muangthanhinius and Bugtile- of Ekgmowechashala is relatively enlarged (Macdonald, mur, the hypoconulids of the lower molars are distinct 1970; Szalay, 1976); canine enlargement is common in and in midline position, while this cusp is more buccally

American Journal of Physical Anthropology 50 J.X. SAMUELS ET AL.

TABLE 2. Estimated body mass (in gm) of Ekgmowechashala specimens using the generalized regressions of body mass as a function of tooth area in Gingerich et al. (1982) Tooth Tooth area Estimated 95% Confidence Specimen position (ln L x W) body mass (g) interval Ekgmowechashala zancanellai JODA 6322 p4 2.425 2191 1997–2404 JODA 6322 m1 2.825 2342 2172–2526 JODA 6322 m3 1.995 866 743–1009 UCMP 128231 P4 2.644 2176 1946–2434 UCMP 128231 M1 2.926 1737 1555–1941 Ekgmowechashala philotau SDSM 5550 p4 2.224 1639 1477–1819 SDSM 62104 p4 2.167 1512 1357–1684 SDSM 5550 m1 2.475 1391 1267–1526 SDSM 62104 m1 2.581 1628 1493–1776 SDSM 55111 m1 2.688 1910 1761–2072 SDSM 5550 m2 2.253 965 866–1075 SDSM 62104 m2 2.346 1089 983–1208 SDSM 55111 m2 2.391 1155 1044–1277 SDSM 62104 m3 2.062 937 808–1087 Tooth areas were calculated using the data in Table 1. located and adjacent to the hypoconid in Ekgmowecha- Ekgmowechashala, based on its low crown height and shala. Muangthanhinius and Bugtilemur have a shallow crenulate enamel (Albright, 2005). However, that tooth, hypoflexid in the lower molars, due to a large laterally identified as a left p4, seems to have a number of char- located cristid obliqua extending from the hypoconid to acteristics quite distinct from Ekgmowechashala. The just distal to the protoconid (Marivaux et al., 2001, tooth bears what seems to be a distinct paraconid mesial 2006). In Ekgmowechashala, the cristid obliqua is later- to and between the protoconid and metaconid, that cusp ally located but very weakly-developed, resulting in a is consistently absent in the p4 of Ekgmowechashala, laterally open talonid basin. though a small paracristid remains. Additionally, in the The body masses of both Ekgmowechashala species Toledo Bend tooth the stylid distobuccal to the protoco- were estimated using generalized regressions of body nid is smaller and more closely appressed to the protoco- mass as a function of tooth area (Table 2; Gingerich nid than in Ekgmowechashala, and it is well separated et al., 1982). The mass of E. zancanellai was consistent from the hypoconid. Additional stylid cusps distal to the across elements, apart from a relatively low estimate entoconid and buccal to the hypoconid of Ekgmowecha- from the m3, which is reduced in that species. Based shala are absent in LSUMG V-2766. Unlike the distinct on the known specimens of E. zancanellai,ithada hypoconid and small postcristid of Ekgmowechashala, mass of approximately 2100 gm, slightly larger than E. the Toledo Bend specimen has a broad crescentic hypoco- philotau.BothEkgmowechashala species were sub- nid/postcristid occupying the distal part of the tooth. stantially larger than Muangthanhinius and Bugtile- The irregular cingulid wrapping around the mesial and mur, which have estimated masses of 365 and 100 g, buccal surfaces of the teeth of Ekgmowechashala are not respectively (Marivaux et al., 2005; Chaimanee et al., present in the Toledo Bend tooth. These differences indi- 2013). These two Asian ekgmowechashalines are cate LSUMG V-2766 is not Ekgmowechashala and fur- similar in size to extant dwarf and mouse lemurs ther study is needed to identify what mammal it (Cheirogaleidae), while Ekgmowechashala was more represents. comparableinsizetolargerspeciesofgalagos(Otole- mur crassicaudatus)andlorises(Nycticebus bengalen- DISCUSSION sis). The mass estimates here are similar to that of Age of Ekgmowechashala from Oregon and the Rose and Rensberger (1983) and indicate Ekgmowecha- Great Plains shala was substantially larger than most North Ameri- can omomyids (Gingerich, 1981), but similar in size to The occurrences of Ekgmowechashala zancanellai are a number of North American and Eurasian adapiforms mapped onto the composite stratigraphic section of the like Cantius, Smilodectes,andAdapis (Gingerich et al., John Day Formation near Sheep Rock in Figure 2. The 1982). holotype of E. zancanellai (JODA 6322) and two addi- While evidence suggests Ekgmowechashala is an ada- tional specimens from Sheep Rock (JODA 16525, 16526) piform, it and the relatively closely related Bugtilemur were recovered from unit H of the Turtle Cove Member and Muangthanhinius remain of questionable affinities of the John Day Formation. The area where these speci- to any of the recognized adapiform primate families mens were collected includes a series of well-exposed without additional material. The distinct hypoconulid in marker beds, with the Picture Gorge Ignimbrite the lower molars of ekgmowechashalines separates them (28.7 6 0.07 Ma) below and Deep Creek Tuff from adapines, cercamoniines, and notharctines, and the (27.89 6 0.57 Ma) above (Albright et al., 2008). That indi- relatively buccal position of the hypoconulid (not cates a well constrained age of between 28.7 and 27.89 twinned to entoconid) separates them from sivaladapids Ma for these occurrences, at the end of the early early (Marivaux et al., 2006). Similarly, the reduced p4 para- Arikareean (Ar1) and near the boundary between the conid and hypoconulid of ekgmowechashalines distin- early and late Oligocene. The magnetostratigraphic guish them from sivaladapines. study by Albright et al. (2008) included samples from An isolated tooth (LSUMG V-2766) from the Toledo the Sheep Rock section, placing unit H within magneto- Bend local Fauna of Texas was previously referred to chron C9r.

American Journal of Physical Anthropology LAST FOSSIL PRIMATE IN NORTH AMERICA 51 The Ekgmowechashala specimen with upper dentition, mengi, Nimravus brachyops, Mesocyon coryphaeus, Nano- UCMP 128231, previously described by Rose and tragulus planiceps, Miohippus annectens, Domnina sp., Rensberger (1983) comes from an unknown stratigraphic and Herpetotherium sp. The Sharps Fauna localities with horizon, but there is sufficient information available to Ekgmowechashala have also yielded Capacikala grada- place it in a stratigraphic context with some confidence. tus, Palaeocastor peninsulatus, Leidymys blacki, Leid- The specimen is described as coming from “sediments ymys woodi, Nanotragulus loomisi, Nimravus brachyops, well above the Picture Gorge Ignimbrite” (Rose and “Mesocyon” temnodon, Miohippus equiceps, Diceratherium Rensberger, 1983). Furthermore, the specimen is indi- armatum,andHerpetotherium youngi (Macdonald, 1970; cated to be associated with specimens of Allomys nitens, Wang, 1994; Wang et al., 1999; Xu, 1996). In addition to above the Meniscomys Local-Range Zone (Rensberger, Ekgmowechashala, the fauna from the Durnal Ranch 1983), and below the Pleurolicus Teilzone (Rensberger, Quarry in Nebraska has yielded specimens of Leidymys 1973) and Entoptychus Concurrent Range Zone blacki, Palaeocastor sp., Miohippus cf. intermedius, Dom- (Rensberger, 1971). While Figure 1 in Rose and nina dakotensis,andHerpetotherium youngi.TheTurtle Rensberger (1983) shows the specimen as occurring above Butte fauna of South Dakota includes occurrences of Pro- the Deep Creek Tuff, there is no indication of whether or tosciurus sp., Capacikala sp., Palaeocastor sp., Pleurolicus not the Deep Creek Tuff is actually present at the locality sp., Paciculus nebraskensis,and“Merycochoerus” super- from which the specimen was collected, or if this place- bus (Martin, 2011). These faunal similarities support con- ment is based upon biostratigraphic interpretation. temporaneous occurrence of Ekgmowechashala in the Recent collections at JODA indicate Allomys nitens Great Plains and Oregon. occurs in units H, K1, and K2 of the Turtle Cove Member of the John Day Formation, as do several species of The origin of Ekgmowechashala Meniscomys. Pleurolicus sulcifrons is restricted to and It is now clear that in the Arikareean, Ekgmowecha- abundant in units K1 and K2 of the Turtle Cove Member, shala appeared simultaneously in Oregon and the Great while Entoptychus species appear in the upper portion of Plains, but its origin is unknown. Muangthanhinius unit K2 and dominate collections from the Kimberly and siami from the latest Eocene of Thailand (34 Ma) and Haystack Valley Member strata (Rensberger, 1971). As Bugtilemur mathesoni from the early Oligocene of Paki- UCMP 128231 was found above the Picture Gorge Ignim- stan (32 Ma) are similar to Ekgmowechashala and brite and with specimens of Allomys nitens, that restricts have recently been placed in the Ekgmowechashalinae its likely occurrence to the upper part of the Turtle Cove (Seiffert, 2007). Similarity to these earlier ekgmowecha- Member, below the Tin Roof Tuff (25.9 6 0.3 Ma, Albright shaline primates from Asia suggests Asian origin for et al., 2008). Additionally, if the specimen was collected this group and an early or late Oligocene immigration to from above the Deep Creek Tuff, then that would suggest North America across Beringia (Seiffert, 2007), which it comes from the uppermost part of the Turtle Cove was open to dispersal through most of the Paleogene Member, in unit K1 or K2, within the late early Arikar- and Neogene (Marincovich and Gladenkov, 1999; Daw- eean (Ar2). Depending on the unit from which UCMP son, 2003). Although the simultaneous occurrence of two 128231 originated, that would place that occurrence species of an enigmatic primate taxon of Asian origin 6 between the Deep Creek Tuff (27.89 6 0.57 Ma) and Bio- million years after the extinction of endemic North tite Tuff (27.14 6 0.13 Ma), or Biotite Tuff and Tin Roof American primates may appear problematic, immigra- Tuff (25.9 6 0.3 Ma) (Albright et al., 2008). tion in the late Eocene or early Oligocene would have Ekgmowechashala philotau was originally described allowed several million years of time for these late Oligo- from the early early Arikareean (Ar1) Wounded Knee/ cene species to diverge. Sharps Fauna (=Sharps Fauna C) from the Sharps For- There are some interesting faunal similarities between mation of South Dakota (Macdonald, 1963, 1970). late Eocene and Oligocene faunas of Asia and Oregon. In Later, Ekgmowechashala was recognized from the addition to Muangthanhinius, the latest Eocene Krabi early early Arikareean (Ar1) Gering Fauna from the Basin of Thailand has yielded specimens of Nimravus cf. Gering Formation of Nebraska (Swisher, 1982). The intermedius and Hoplophoneus sp. (Peigne et al. 2000). upper Sharps Formation and Gering Formation span Nimravus is also known from the early Oligocene of the same magnetostratigraphic interval as Turtle Cove Mongolia (de Bonis, 1981; Lange-Badre and Dashzeveg, unit H, magnetochron C9r (Tedford et al., 1996). Ekg- 1989; Dashzeveg, 1996) and Nimravus brachyops is the mowechashala sp. has also been noted from the late most common nimravid in the Turtle Cove fauna, rang- early Arikareean (Ar2) aged Turtle Butte Formation of ing from unit C through K1. Several species of Palaeo- South Dakota (Schumaker, 2008; Martin, 2011). These gale occur in the early Oligocene of Mongolia; Palaeogale records confirm the presence of Ekgmowechashala in sp. is known from units E through H of the Turtle Cove these two regions at the same time, the early Arikar- Member in Oregon. Also in Oregon in the early and late eean (Ar1-Ar2). Oligocene (Ar1) is a new genus of beaver (Korth and A number of other taxa from these Great Plains local- Samuels, In Press) that is very similar to Propalaeocas- ities are the same species or congeneric with taxa known tor from the early Oligocene of Europe and Asia, includ- from the upper Turtle Cove fauna. Along with Ekgmowe- ing P. irtyshensis from the early Oligocene of China (Wu chashala, other taxa known from unit H (Ar1) in the et al., 2004). These records suggest a number of taxa dis- vicinity of Sheep Rock include the following: Capacikala persed between Asia and western North America in the gradatus, Leidymys nematodon, Leidymys parvus, Nano- late Eocene or Oligocene. tragulus planiceps, Miohippus annectens,andDicerathe- rium armatum. Among the most common taxa found in Paleoecology of Ekgmowechashala units K1 and K2 (Ar2) in the vicinity of Sheep Rock are Capacikala gradatus, Palaeocastor peninsulatus, Pleuroli- The upper and lower molars of Ekgmowechashala are cus sulcifrons, Leidymys nematodon, Paciculus insolitus, characterized by having bulbous cusps with poorly and Promerycochoerus superbus, as well as Protosciurus defined crests, including the cristid obliqua, and large

American Journal of Physical Anthropology 52 J.X. SAMUELS ET AL. basins. Several studies of primates have found the and differences from the previously described E. philo- molars of frugivorous taxa display features that facili- tau indicate the presence of a distinct new species. Com- tate crushing, including transversely oriented protocris- parison of the known material of Ekgmowechashala tid, buccally located cristid obliqua, and larger talonid allows a revised assessment of its relationship to other basins than faunivorous taxa (Strait, 1993; White, 2009). primates, suggesting placement among the adapiforms. The weakly developed crests and large basins of Ekgmo- Ekgmowechashala and two recently described ekgmowe- wechashala suggest that crushing, rather than shearing, chashaline primates from Asia share a double-rooted p2, was the primary function of the teeth and that its diet molariform p4, mesiodistally compressed molar trigo- was primarily frugivorous, as was suggested by Szalay nids, reduced or absent molar paraconids, distinct molar (1976) and Rose and Rensberger (1983). Without more hypoconulids, and crenulated enamel, among other fea- complete cranial and postcranial material, little else can tures. Similarity of Ekgmowechashala to these two pri- be said regarding the ecology of Ekgmowechashala. mates from the late Eocene and early Oligocene of Asia suggest ekgmowechashalines originated in Asia and dis- Late Eocene and Oligocene paleoenvironments persed to North America in the Oligocene, 6 million in Oregon years after other primates on the continent became extinct. These records also help refine the age range of Fossil leaf beds near the Eocene-Oligocene transition this primate in North America, which was limited to the have been recognized at many localities across central early Arikareean (Ar1–Ar2). Combined, the Oregon and and eastern Oregon, documenting the ecosystem changes Great Plains records indicate the last non-human prima- associated with changes toward cooler and more seasonal tes in North America vanished in the late Oligocene, climates in the early Oligocene (Meyer and Manchester, about 26 million years ago. 1997; Manchester, 2000; Dillhoff et al., 2009). The Bridge Creek Flora assemblage lies near the Eocene–Oligocene ACKNOWLEDGMENTS boundary itself, dated between 32.99 6 0.11 Ma and 32.66 6 0.03 Ma (Retallack et al., 2000). This assemblage Constructive comments by Christopher Beard (Univer- documents dramatic floral shifts that coincide with sity of Kansas), Erik Seiffert (Stony Brook University), regional and global climate changes, characterized by and the editors substantially improved this manuscript. gradual cooling, increased aridity, and more seasonal pre- The authors thank Laurent Marivaux (Institut des Scien- cipitation through time (Meyer and Manchester, 1997; ces de l’Evolution de Montpellier) for providing helpful Manchester, 2000; Retallack et al., 2000; Retallack, 2004, references. And many thanks go to Keila Bredehoeft for 2007; Zachos et al., 2001, 2008; Dillhoff et al., 2009). providing the original artwork in this article. Paleosol and faunal evidence from the Turtle Cove Member also document significant environmental changes LITERATURE CITED in the Oligocene of Oregon. Through the Arikareean, climate conditions became cooler and drier, woodland habi- Albright LB. 2005. Ekgmowechashala (Mammalia, ?Primates) tats began to open up, and bunch grasses and shrubs from the Gulf Coastal Plain. Bull Florida Museum Nat Hist became more common (Retallack et al., 2000; Retallack, 45:355–361. Albright LB, Woodburne MO, Fremd TJ, Swisher CC, 2004, 2007). This time period includes the appearance of MacFadden BJ, Scott GR. 2008. Revised chronostratigraphy taxa with higher crowned teeth, as well as the region’s and biostratigraphy of the John Day Formation (Turtle Cove first appearance of burrowing mammals (palaeocastorine and Kimberly Members), Oregon, with implications for beavers, geomyids, and proscalopids) and cursorial mam- updated calibration of the Arikareean North American Land mals (Archaeolagus ennisianus) (Albright et al., 2008; Mammal Age. J Geol 116:211–237. Korth and Samuels, in press). Occurrence of these open Anemone RL, Skinner MM, Dirks S. 2012. Are there two dis- habitat adapted taxa with clearly arboreal species, like tinct types of hypocone in Eocene primates? The Protosciurus and Miosciurus, support reconstruction of “pseudohypocone” of nartharctines revisited. Palaeontol Elec- Turtle Cove as having a mosaic open woodland environ- tron 15:13. ment in the Oligocene. This contrasts strongly with the Beard KC. 2008. The oldest North American primate and mammalian biogeography during the Paleocene-Eocene thermal habitats inhabited by earlier North American primates, maximum. Proc Natl Acad Sci USA 105:3815–3818. specifically relatively warm, closed subtropical forests Bonis LD. 1981. Contributions a l’etude du genre Palaeogale (Gunnell et al., 2008; Wing, 1998). meyer (mammalia, carnivora). Ann Paleontol (Vertebres) 67: The extinction of Ekgmowechashala is likely attribut- 37–56. able to continued cooling and drying through the late Bown TM, Rose KD. 1987. Patterns of dental evolution in early Oligocene. The early late Arikareean (Ar3) aged Kim- eocene anaptomorphine primates (Omomyidae) from the Big- berly Member overlying Turtle Cove is dominated by horn Basin, Wyoming. Paleontol Memoir 23:1–162. open habitat adapted mammals, specifically the entopty- Chaimanee Y, Chavasseau O, Lazzari C, Euriat A, Jaeger JJ. chine gopher Entoptychus (Rensberger, 1971), and paleo- 2013. A new late Eocene primate from the Krabi Basin (Thai- sols from these strata suggest much more arid and open land) and the diversity of Palaeogene anthropoids in south- east Asia. Proc Biol Sci 280:20132268. Available at: http://dx. environments (Retallack et al., 2000; Retallack, 2007). doi.org/10.1098/rspb.2013.2268 As the woodland habitats inhabited by Ekgmowecha- Dashzeveg D. 1996. Some carnivorous mammals from the Pale- shala declined, this last fossil primate in North America ogene of the eastern Gobi Desert, Mongolia, and the applica- vanished with them. Primates were not seen again in tion of Oligocene carnivores to stratigraphic correlation. Am North America until the arrival of humans more than 25 Mus Novit 3179:1–14. million years later. Dawson MR. 2003. Paleogene rodents of Eurasia. In: Reumer JWF, Wessels W, editors. Distribution and migration of Terti- CONCLUSIONS ary mammals in Eurasia, a volume in honor of Hans de Bruijn. Deinsea 10:97–126. The new material of Ekgmowechashala described here Dillhoff RM, Dillhoff TA, Dunn RE, Myers JA, Stromberg€ CAE. helps to confirm the presence of this primate in Oregon, 2009. Cenozoic paleobotany of the John Day Basin, central

American Journal of Physical Anthropology LAST FOSSIL PRIMATE IN NORTH AMERICA 53

Oregon. In: O’Connor JE, Dorsey RJ, Madin IP, editors. Volca- McKenna MC. 1990. Plagiomenids (Mammalia:? Dermoptera) noes to Vineyards: geologic field trips through the dynamic from the Oligocene of Oregon, Montana, and South Dakota, landscape of the Pacific Northwest. Geol Soc Am Field Guide and middle Eocene of northwestern Wyoming. In: Bown TM, 15:135–164. Rose KD, editors. Dawn of the age of mammals in the north- Fisher RV, Rensberger JM. 1972. Physical stratigraphy of the ern part of the Rocky Mountains Interior, North America. John Day Formation, central Oregon. Univ California Publ Geol Soc Am Spl Pap 243:211–234. Geol Sci 101:1–33. Merriam JC. 1901. A contribution to the geology of the John Fremd TJ. 2010. Guidebook: SVP Field Symposium 2010, John Day Basin. Univ California Publ Geol Sci 2:269–314. Day Basin Field Conference. Soc Vertebr Paleontol 1–153. Meyer HW, Manchester SR. 1997. The Oligocene Bridge Creek Fremd TJ, Bestland EA, Retallack GJ. 1994. John Day Basin flora of the John Day Formation, Oregon. Univ California field trip guide and road log. Soc Vertebr Paleontol 1–80. Publ Geol Sci 141:1–195. Gingerich PD. 1981. Early Cenozoic Omomyidae and the evolu- Ni X, Meng J, Beard KC, Gebo DL, Wang Y, Li C. 2010. A new tionary history of tarsiiform primates. J Hum Evol 10:345– tarkadectine primate from the Eocene of inner Mongolia, 374. China: phylogenetic and biogeographic implications. Proc Biol Gingerich PD. 1995. Sexual dimorphism in earliest Eocene Can- Sci 277:247–256. tius torresi (Mammalia, Primates, Adapoidea). Contributions Ostrander G. 1987. The early Oligocene (Chadronian) Raben from the Museum of Paleontology. Univ Michigan 29:185–199. Ranch local fauna, Northwest Nebraska: Marsupialia, Insec- Gingerich PD, Smith BH, Rosenberg K. 1982. Allometric scaling tivora, Dermoptera, Chiroptera, and Primates. Dakoterra 3: in the dentition of primates and prediction of body weight 92–104. from tooth size in fossils. Am J Phys Anthropol 58:81–100. Peigne S, Chaimanee Y, Jaeger J-J, Suteethorn V, Ducrocq S. Gunnell GF, Rose KD, Rasmussen DT. 2008. Euprimates. In: 2000. Eocene nimravid carnivorans from Thailand. J Vertebr Janis CM, Gunnell GF, Uhen MD, editors. Evolution of Terti- Paleontol 20:157–163. ary mammals of North America, Vol.2. New York: Cambridge Rensberger JM. 1971. Entoptychine pocket gophers (Mammalia, University Press. p 239–261. Geomyoidea) of the early Miocene John Day Formation, Ore- Hay RL. 1963. Stratigraphy and zeolitic diagenesis of the John gon. Univ California Publ Geol Sci 90:1–209. Day Formation of Oregon. Univ California Publ Geol Sci 42: Rensberger JM. 1973. Pleurolicine rodents (Geomyoidea) of the 199–262. John Day Formation, Oregon and their relationships to taxa from the early and middle Miocene, South Dakota. Univ Cali- Hunt RM, Stepleton E. 2004. Geology and paleontology of the fornia Publ Geol Sci 102:1–95. upper John Day beds, John Day River Valley, Oregon: lithos- Rensberger JM. 1983. Succession of meniscomyine and allo- tratigraphic and biochronologic revision in the Haystack Val- myine rodents (Aplodontidae) in the Oligo-Miocene John Day ley and Kimberly areas (Kimberly and Mt. Misery Formation, Oregon. Univ California Publ Geol Sci 124:1–157. quadrangles). Bull Am Mus Nat His 282:1–90. Retallack GJ. 2004. Late Oligocene bunch grassland and early Janis CM. 1993. Tertiary mammalian evolution in the context Miocene sod grassland paleosols from central Oregon, USA. of changing climates, vegetation, and tectonic events. Annu Palaeogeogr Palaeoclimatol Palaeoecol 207:203–237. Rev Ecol Syst 24:467–500. Retallack GJ. 2007. Cenozoic paleoclimate on land in North Korth WW, Samuels JX. 2015. New rodent material from the America. J Geol 115:271–294. John Day Formation (Arikareean, Middle Oligocene to Early Retallack GJ, Bestland EA, Fremd TJ. 2000. Eocene and Oligo- Miocene) of Oregon. Annals of Carnegie Museum, 83:19–84. cene paleosols of central Oregon. Geol Soc Am Spl Pap 344:1– Lange-Badre B, Dashzeveg D. 1989. On some Oligocene carniv- 196. orous mammals from central Asia. Acta Palaeontol Pol 34: Robinson PT, Brem GF, McKee EH. 1984. John Day Formation 125–148. of Oregon: a distal record of early Cascade volcanism. Geology Macdonald JR. 1963. The Miocene faunas from the Wounded 12:229–232. Knee area of western South Dakota. Bull Am Mus Nat Hist Rose KD. 2006. The beginning of the age of mammals. Balti- 125:139–238. more: The Johns Hopkins University Press. Macdonald JR. 1970. Review of the Miocene Wounded Knee fau- Rose KD, Rensberger JM. 1983. Upper dentition of Ekgmowe- nas of southwestern South Dakota. Bull Los Angeles County chashala (omomyid primate) from the John Day Formation, Museum Nat Hist 8:1–82. Oligo-Miocene of Oregon. Folia Primatol 41:102–111. Manchester SR. 2000. Late Eocene fossil plants of the John Day Schumaker KK. 2008. Paleontology, biostratigraphy, and geol- Formation, Wheeler County, Oregon. Oregon Geol 62:51–63. ogy of the Turtle Butte Formation (latest Oligocene-early Mio- Marincovich L, Gladenkov AY. 1999. Evidence for an early cene), Tripp County, South-Central South Dakota. M.S. opening of the Bering Strait. Nature 397:149–151. Thesis, South Dakota School of Mines and Technology. p. 1– Marivaux L, Antoine PO, Baqri SRH, Benammi M, Chaimanee 229. Y, Crochet JY, de Franceschi D, Iqbal N, Jaeger J-J, Metais Seiffert ER. 2007. Evolution and extinction of Afro-Arabian pri- G, Roohi G, Welcomme JL. 2005. Anthropoid primates from mates near the Eocene–Oligocene boundary. Folia Primatol the Oligocene of Pakistan (Bugti Hills): data on early anthro- 78:314–327. poid evolution and biogeography. Proc Natl Acad Sci USA Simons EL. 1962. A new eocene primate genus, Cantius, and a 102:8436–8441. revision of some allied European lemuroids. Bull Br Mus 7:1– Marivaux L, Chaimanee Y, Tafforeau P, Jaeger J-J. 2006. New 36. strepsirrhine primate from the late Eocene of peninsular Smith T, Rose KD, Gingerich PD. 2006. Rapid Asia–Europe– Thailand (Krabi Basin). Am J Phys Anthropol 130:425–434. North America geographic dispersal of earliest Eocene pri- Marivaux L, Welcomme JL, Antoine PO, Metais G, Baloch IM, mate Teilhardina during the Paleocene–Eocene thermal max- Benammi M, Chaimanee Y, Ducrocq S, Jaeger J-J. 2001. A imum. Proc Natl Acad Sci USA 103:11223–11227. fossil lemur from the Oligocene of Pakistan. Science 294:587– Storer JE. 1990. Primates of the Lac Pelletier lower fauna 591. (Eocene: Duchesnean), Saskatchewan. Can J Earth Sci 27: Martin JE. 2011. The Rosebud problem revisited. Proc South 520–524. Dakota Acad Sci 90:37–50. Strait SG. 1993. Differences in occlusal morphology and molar McClaughry JD, Ferns ML, Streck MJ, Patridge KA, Gordon size in frugivores and faunivores. J Hum Evol 25:471–484. CL. 2009. Paleogene calderas of central and eastern Oregon: Swisher CC. 1982. Stratigraphy and biostratigraphy of the east- eruptive sources of widespread tuffs in the John Day and ern portion of the Wildcat Ridge, Western Nebraska. M.S. Clarno Formations. In: O’Connor JE, Dorsey RJ, Madin IP, Thesis, University of Nebraska. p. 1–172. editors. Volcanoes to Vineyards: geologic field trips through Szalay FS. 1976. Systematics of the Omomyidae (Tarsiiformes, the dynamic landscape of the Pacific Northwest. Geol Soc Am Primates) taxonomy, phylogeny, and adaptations. Bull Am Field Guide 15:407–434. Mus Nat Hist 156:157–449.

American Journal of Physical Anthropology 54 J.X. SAMUELS ET AL.

Szalay FS, Delson E. 1979. Evolutionary history of the prima- White J. 2009. Geometric morphometric investigation of molar tes. New York: Academic Press. shape diversity in modern lemurs and lorises. Anat Rec 292: Szalay FS, Lucas SG. 1996. The postcranial morphology of pale- 701–719. ocene Chriacus and Mixodectes and the phylogenetic relation- Wilson JA. 1966. A new primate from the earliest Oligocene, ships of archontan mammals. New Mexico Museum Nat Hist West Texas, preliminary report. Folia Primatol 4:227–248. Sci Bull 7:1–46. Wing SL. 1998. Tertiary vegetation of North America as a con- Tedford RH, Swinehart JB, Swisher CC, Prothero DR, King text for mammalian evolution. In: Janis CM, Scott KM, SA, Tierney TE. 1996. The Whitneyan-Arikareean transi- Jacobs L, editors. Evolution of tertiary mammals of North tion in the High Plains. In: Prothero DR, Emery RJ, edi- America. New York: Cambridge University Press. p 37–65. tors. The terrestrial Eocene-Oligocene transition in North Wu WJ, Meng J, Ye J, Ni X. 2004. Propalaeocastor (Rodentia, America. New York: Cambridge University Press. p 312– Mammalia) from the early Oligocene of Burqin Basin, Xin- 324. jiang. Am Mus Novit 3461:1–16. Townsend KEB, Rasmussen DT, Murphey PC, Evanoff E. 2010. Xu XF. 1996. Castoridae. In: Prothero DR, Emery RJ, editors. Middle Eocene habitat shifts in the North American western The terrestrial Eocene-Oligocene transition in North America. interior: a case study. Palaeogeogr Palaeoclimatol Palaeoecol New York: Cambridge University Press. p. 417–432. 297:144–158. Zachos JC, Dickens GR, Zeebe RE. 2008. An early Cenozoic per- Wang X. 1994. Phylogenetic systematics of the Hesperocyoninae spective on greenhouse warming and carbon-cycle dynamics. (Carnivora: Canidae). Bull Am Mus Nat Hist 221:1–207. Nature 451:279–283. Wang X, Tedford RH, Taylor BE. 1999. Phylogenetic systematics Zachos JC, Pagani M, Sloan L, Thomas E, Billups K. 2001. of the Borophaginae (Carnivora: Canidae). Bull Am Mus Nat Trends, rhythms, and aberrations in global climate 65 Ma to Hist 243:1–391. present. Science 292:686–693.

American Journal of Physical Anthropology